Due to SPAM attacks, new members must be approved before posting.  Please email when registering and your account will be approved.

Main Menu

MTL to NAVD88 Correction

Started by Jonathan S. Clough, March 29, 2010, 11:52:41 AM

Previous topic - Next topic

Jonathan S. Clough

The question was asked:
QuoteFrom: rloiselle   on: March 26, 2010, 02:49:06 PM
Examining our elevation data for conversion errors in the data processing of date and datum conversions, stated on pg 9 of the manual, has lead us to question whether we should be adding or subtracting the difference between MTL and NAVD88 from the DEM elevations.  In our mind, because we want to set the NAVD88 elevations to an MTL datum, and our NAVD88 elevations are located below our MTL, we feel like we should be adding that difference to bring our elevations up to the MTL level.  My understanding is that the East Coast MTL is lower than their NAVD88 and so subtracting that difference would bring their elevations down to MTL?

The NAVD to MTL correction (or more generally [any vertical datum] to MTL correction) is utilized to convert the elevation data maps from a fixed vertical datum to a datum based on the Mean Tide Level of the site.  It is defined as MTL minus NAVD88 based on the following type of data from a NOAA gage:

( e.g. )

          MTL           8.235  Mean Tide Level
          MSL           8.224  Mean Sea Level
          MLW           7.440  Mean Low Water
          MLLW          7.025  Mean Lower-Low Water
          LWI           1.27   Greenwich Low  Water Interval (in Hours)
          NAVD          7.176  North American Vertical Datum

You can see from these results that, at this site, an elevation of 1.0 meters in NAVD88 units will actually be located below MTL.

In this case, the MTL minus NAVD88 will be roughly 1.06 meters.  As a general rule, if your MTL-NAVD parameter goes up then elevations go down.

If you are using the NOAA VDATUM product: ( you output "height" in "meters" and then use "MTL" as the input and "NAVD88" as the output.  That should give you the number that you are looking for with the proper units and sign.


We did a SLAMM run and calculated the NAVD_correction as MTL-NAVD88 = -0.067.  This gave us a poor T0 result with lots of changes.  We recalculated with vdatum and got a result of -0.5066, which gave a a much improved T0 result.  However, we had used "MLW" as the Input Vertical Datum to get that result.  When we use "MTL" as the Input Vertical Datum, we get -0.0452, which is a poor T0 result.  Any ideas about why this is happening or what we should look into? 

Jonathan S. Clough

You will need to use MTL as the input vertical datum.  The proper correction is obviously much closer to -0.05 than -0.5 meters, I'm afraid.

As you know, the discrepancy in "time-zero" results gives you a partial sense of the uncertainty of the model, even in the current condition.  There will almost always be some discrepancy due to horizontal offsets between the land-cover and elevation data sets, general data uncertainty, or other local conditions that make a portion of your site not conform perfectly to the "conceptual model." 

However, when looking for larger problems:

  • first run an elevation analysis and see which categories are not matching the conceptual model well (comparing 5th and 95th percentiles to model expectations.).  Read the Users' Manual section on Elevation Analysis
  • don't forget the caveat about tidal swamp and tidal fresh marsh presented in the Technical Documentation, page 16
  • ensure that your tidal ranges are properly parameterized spatially throughout the map (tidal ranges will often decrease up-estuary)
  • ensure that the converting land is not somehow protected by dikes or levees.
  • is there some other reason that local conditions are causing this early conversion?
  • you can adjust the SLAMM conceptual model for your site if you believe it's warranted.

That's off the top of my head, I'll try to add to this list as I think of more.

Good luck!  -- Jonathan


I am using the VDatum to do vertical datum correction, the test data is NED data ,Every time i try to convert an  elevation point in geographic coordinate the convertion result of the Latitude/Longitude  have no changes, but the height is always -999999.0.Would you tell me what is wrong in the process.

thank you.

Jonathan S. Clough

VDATUM is not very user friendly.  You need to have unzipped and selected the correct geographic region data file and have the lat and long coords. precise.  Then it will not return any information over dry land so you'll need some sort of interpolation procedure for that portion of the study area.  -999999.0 is the VDATUM code for No-Data.

That's all I can tell you.  Other than that you'll have to communicate with NOAA to get support for their VDATUM product.  -- Jonathan


The VDatum's tidal transformations are valid in the water areas only, i.e. all points in lands, islands and points that are out of our tidal transformation grids will have height results as -999999.0,when i change the point to water area,it makes sense,the point conversion result turn out to be -0.036.